J.Org. Chem., Vol. 43, No. 22, 1978
Notes Reaction of 2 with morpholine: 2.85 g (42%);mp 64-65 OC; IR (cc14) 3060, 2960, 2850, 1580, 1460, 1440,1250, 1130, 1115, 1010,905,860, 690 cm-'; NMR (CDC13) 6 1.3 (d, 3 H, J = 7 Hz, CH3), 2.45 (m, 4 H), 3.4 (dq, 1H, 5 2 , 3 = 2.5 Hz, C3-H),3.6 (m, 4 H), 4.65 (d, 1H, C2-H),7.1 (m, 4 H, aromatic H); mass spectrum, mle (relative intensity) 235 (M+, 36), 202 (4), 149 (loo), 148 (92), 147 (60), 135 (281,134 (641,114 (16), 115 (24), 105 (16), 103 (16), 100 (34), 91 (251, 77 (28). Anal. Calcd for C13Hl;NOS: C, 66.34; H, 7.28; N, 5.95; 0,6.80; S, 13.62. Found: C, 66.35; H, 7.16; N, 6.16; 0,6.99; S,13.74. Reaction of Piperidine and Benzo[ blthiophene by Means of n-Butyllithium. n-Butyllithium (0.15 M, 20% solution in hexane) was added to 60 mL of piperidine under nitrogen. The temperature of the mixture was maintained at 40 OC, and a solution of 4 g (0.03 M) of benzo[b]thiophene in 10 mL of amine was added. Reaction and
isolation were performed as previously described. 2-Piperidino2,3-dihydrobenzo[b]thiophene(4)was purified by chromatography, 3 g (46%). Reaction of Morpholine and Benzo[ blthiophene by Means of Sodium Hydride. A mixture of 60 mL of morpholine and 3.6 g (0.15 M) of NaH was refluxed under nitrogen until the evolution of hydrogen ceased and was cooled to 40 "C. The addition of benzo[b]-
thiophene and the reaction procedure were as previously described. 2-Morpholino-2,3-dihydrobenzo[b]thiophene ( 5 ) was recrystallized from heptane-toluene, 2.7 g (41%). 2-Morpholinobenzo[ blthiophene (12). A 1.1-g (0.005 mol) amount of 5 and 0.16 g (0.005 mol) of sulfur were heated at 250 "C until the evolution of H2S ceased (5 min). The reaction mixture was taken into benzene and decolorized with Norit. Evaporation of benzene and recrystallization from toluene-heptane gave a colorless solid: 0.65 g (60%);mp 179-180 "C; IR (CC14) 3060,2960,2900,2855,2815, 1530,1440,1120,1030,930,900,870,650 cm-l; NMR (CCla)& 3.2 (m, 4 H), 3.9 (m, 4 H I ,6.2 (s. 1 H, C3-H),7.3 (m, 4 H, aromatic H);mass spectrum, rnle (relative intensity) 219 (M+, loo), 204 (6), 162 (23), 161 (93), 160 (381, 147 (141, 135 (81, 134 (261, 133 (9), 89 (201, 80 (12). Anal. Calcd for C12H13NOS: C, 65.72; H, 5.97; N, 6.39; 0, 7.30; S, 14.62. Found: C, 65.71; H, 5.95; N, 6.35; 0, 7.55; S, 14.58. 2-Morpholino-3-methylbenzo[blthiophene (13). From aromatization of 100 mg of 9 with 15 mg of sulfur: 65 mg (65%);mp 79-80 "C; IR (CC14) 3060, 2960. 2900, 2855, 2820, 1575, 1435, 1190, 1120, 1045, 1015, 980, 880 cm-I; NMR (Cc14) 6 2.3 (s, 3 H, CH3), 3.0 (m, 4 H), 3.9 (m, 4 H), 7.4 (m, 4 H, aromatic H); mass spectrum, rnle (relative intensity) 234 (181,233 (M+, 1001,232 (271,218 (51, 188 (51,176 i l l ) , 175 (50), 174 (58), 173 (16), 161 (12), 160 (301,159 (lo),147 (30), 134 (11). Anal. Calcd for ClnHlsNOS: mol wt 233.0874. Found (high-resolution mass spectrum): mol w t 233.0876.
Facile S y n t h e s i s of 2-Substituted Imidazoles Kenneth L. Kirk Laboratory of Chemistry, National Institute of Arthritis, Metabolism, and Digestive Diseases, National Institutes of Health, Bethesda, Maryland 20014 Received June 13,1978 Continuing studies in the biochemistry and pharmacology of ring-fluorinated imidazoles have revealed striking differences in behavior between 2- and 4-flUOrO isomers in each series.1 For example, 2-fluorohistidine displays a wide range of biological activities2 while 4-fluorohistidine shows little or no activity in t h e same systems. As part of our efforts to elucidate the causes of these differences, we wished to extend our testing t o isomer pairs of t h e other haloimidazoles-particularly t h e halohistidines. T h e 4 (or 5)-halo derivatives can be obtained by direct electrophilic substitution,3 but no methods are available for preparation of the 2-halo isomers. While 2fluoro-4 and 2-chloroimidazoles5 have been prepared by photochemical decomposition of 2-diazoniumimidazoles, the method fails for bromine or iodine, and there exists no obvious procedure for the introduction of the latter halogem6 We have now developed a general synthesis, not only for 2-haloimidazoles, b u t for a variety of other 2-substituted imidazoles. In 1-alkyl or 1-arylimidazoles (methyl, benzyl, phenyl), H-2 is t h e most acidic hydrogen and a carbanion is readily generated at C-2 by reaction with n-butyllithium; this carbanion has been used for addition to carbonyl groups7 as well as t o other electrophilic reagents8 Unfortunately, the 1-substituent is not easily removed from the product in these cases. N Benzylimidazole can be debenzylated with sodium in liquid ammonia? b u t bromine or iodine a t C-2 undoubtedly would be removed a t t h e same time. We have found t h a t 1-tritylimidazolelO also forms a carbanion (1) with n-butyllithium,
1
Registry No.--l,95-15-8; 2,1455-18-1;cis-7,66902-28-1; trans-7, 66902-27-0;cis- 8,66902-26-9; trans- 8,66902-22-5; cis- 9,66902-21-4; trans- 9,66902-20-3: 12, 18774-55-5; 13,66902-25-8.
References a n d Notes (1) (a)R. Wegler and G. Pieper, Chem. Ber., 83, 1 (1950); (b) H. Bestian, Justus Liebigs Ann. Chem., 566, 210 (1950). (2) R. D. Ciosson, ,I. P. Napoiitano, G. G. Ecke, and A. J. Koika, J. Org. Chem., 22, 646 (1957) (3) (a)L. E. Harris, D. V. Hertzier, 0. C. Dermer, and E. J. Eisenbraun, J. Org. Chem., 37,3039 (1972); (b) E. J. Eisenbraun, R. C. Bansal, D. V. Hertzier, W. P. Duncan, P. W. K . Flanagan, and C. M. Hamming, hid., 35, 1265 (1970); (c)R. C. Bansai, E. J. Eisenbraun, and P. W. Flanagan, J. Am. Chem. Soc., 88, 1837 (1966). (4) R. A. Benkeser, R. K. Agnihotri. M. L. Burrous, E. M.Kaiser, J. M.Mailan, and P. W. Ryan, J, Org. Chem., 29, 1313 (1964). (5) Birch-type reduction of 1 has been effected by the milder reducing system sodium-ethanol-ammonia to give 2-ethylthiophenol: W. Huckel and I. Nabih, Chem. Ber., 89, 2115 (1956). (6) When the reaction is performed with n-butyliithium and a primary amine, no reduction products are formed but the yield of adduct remains disap-
pointingly low.
(7) Cis and trans isomers of adducts 7,8, and 0 cannot be separated,even at
an analytical scale. The assumption that the major product is the trans isomer is based on literature data assuming that the chemical shift of H2 should be higher in the cis isomer than in the trans isomer and that J2.3 is higher when H2 and H3 are cis than when they are trans8 (8) (a)L. M. Jackman and S. Sternhell, "Application of NMR Spectroscopy In Organic Chemistry", 2nd ed, Pergamon Press, Oxford, 1969, p 233; (b) J. J. Christol, T. W. Russel, J. R. Mohrlg, end D. C. Plorde, J. Org. Chem.,
31, 581 (1966). (4)The yield is less then 1% based on consumed benzo[b]thiophene. (10) (a)K. R. Brower and E. D. Amstutr, J. Org. Chem., 18, 41 1 (1954); (b) 8 . lddon and R . M. Scrowston, Adv. Heterocycl. Chem., 11, 177 (1970). (I 1) E. G. G. Werner, Red. Trav. Chlm. Pays-Bas, 88, 509 (1949).
4381
2
3
t h a t t h e carbanion reacts readily with various electrophiles t o form 1-trityl-2-X-imidazoles (Z), and that the trityl group is easily removed by mild acid hydrolysis to give 2-X-imidazoles (3). Tables I and I1 describe compounds prepared by this general method. Yields of 2 are consistently high,ll except where X is halogen. Attempts to improve yields in the halogenation steps by variation in conditions or source of halogen were unsuccessful. Unreacted tritylimidazole accounted for most of t h e material loss in these cases. T h e presence of a single imidazole proton resonance in the NMR spectrum of each 3 supports assignment of the substituent to the 2-position. For 3c, 3d, a n d 3g, structural assignments were confirmed by comparison with authentic samples. In no case was there formed a detectable quantity of the isomeric 4(5)-X-imidazole, based on NMR and chromatographic evidence. T h e preparation of 2-aminoimidazole through t h e phenyltriazene (2g), based on a procedure for t h e preparation of l-alkyl-2-aminoimidazoles,~2 has special significance in t h a t it allows a nonreductive introduction of the 2-amino function into a preformed imidazole ring. (In our hands, the catalytic reduction of 2-arylazo-4-X-imidazoles often results in simultaneous loss of the 4-X substituent.5) Consistent with the results of others,lZ our attempts t o aminate 1 with methoxyamine,13 O-mesitylenesulfonylhydroxylamine,*4 or 0-2,4dinitrophenylhydroxylaminel5were unsuccessful.
This article not subject t o U.S. Copyright. Published 1978 by t h e American Chemical Society
4382 J.Org. C'hern., Vol. 43, No. 22, 1978
Notes
Table I. Products of the Reaction of 1-Trityl-2-lithioimidazolewith Electrophilic Agents registry no.
electrophilic agent
cpd"
registry no.
yield,
X
%
mp, "C
NMR, ppm ( J , Hz)
67478-46-0 2a N-iodosuccinimide
516-121
I
40
170-172
6.81 (d) (J = 1.5), 6.98 (d) ( J = 1.5), 7.05-7.38 (m)
2a 1 2 c 67478-47-1 2b N-bromosuccinimidec
7553-56-2 128-08-5
I Br
41 35
208-209
67478-48-2
2~ N-chlorosuccinimidec
128-09-6
c1
99 16681-56-4 sublimation 7.07 (s) A 98 166-167' 3c c1 16265-04-6 130-133/ sublimation 2.29 (s), 6.89 (s) 3d CH3 B 97 693-98-1 178-179 ethanol 1.30 (t) ( J = 7), 3.35 (9) ( J = 7),7.32 (s) 3e C02C2H5 A >99 33543-78-1 190-196g ethyl acetate 7.41 (s), 9.63 (s) 3f CHO C 99 10111-08-7 144-146h ethanol 6.88 (s) 3g NHZ.HC1 D 73 52737-40-3 All new compounds had satisfactory elemental analyses. Identity of all compounds was checked by mass spectrometry. (A) 1 mmol refluxed 30 min in 5 mL of 5% acetic acid in methanol; (B) 1 mmol refluxed 4 h in 1mL of 1 N HC1 and 0.5 mL of ethanol; (C) 1 mmol refluxed for 1 h in 5 mL of 5% acetic acid in methanol; (D) 1 mmol refluxed for 3 h in 10 mL of methanol and 0.2 mL of concentrated HC1. < NMR spectra were measured on a JEOL Model FX 100 spectrometer. d Lit. mp 207 "C: H. King and W. 0. Murch, J . Chem. Soc., 123,621 (1923). e Lit.6 mp 165-166 "C. f Lit. mp 137 'C: 0. Wallach, Justus Liebigs Ann. Chem., 214,257 (1882). g Lit. mp 204 "C: H. Shubert and H.-D. Rudolf, Angew. Chem. Znt. Edit. Engl., 5,674 (1966). Lit. mp 152 "C: R. G. Fargher and F. L. Pyman, J. Chem. SOC.,115, 217 (1919). T h e procedure described in this report is now being applied t o t h e preparation of 2,4-disubstituted imidazoles and t o 2substituted histidines and histamines. The variety of X groups introduced is also being expanded. E x p e r i m e n t a l Section Preparation of 1-Trityl-2-X-imidazoles (2). The preparation of 1-trityl-2-iodoimidazole(2a) illustrates the general procedure. 1-Trityl-24ithioimidazole(I) was prepared by the addition of 1.5 mL of 1.6 M n-butyllithium in hexane (Aldrich) to a solution of 620 mg (2 mmol) of 1-tritylimidazole in 25 mL of tetrahydrofuran (freshly distilled from lithium aluminum hydride) at 0 "C under a nitrogen atmosphere. The solution, which gradually turned red, was stirred at room temperature for 1.5 h, was then cooled to 0 "C, and 508 mg (2 mmol) of iodine in 5 mL of tetrahydrofuran was added dropwise over 5 rnin. After an additional 10 min at 0 "C, the reaction mixture was poured into 25 mL of water. After concentration of the solution by rotary evaporation, ether extraction, and silica gel chromatography (1:l ether-petroleum ether), 2a was obtained in 40% yield. Products described in Table I were prepared from 2 mmol of tritylimidazole. No problems are encountered when the reaction is carried out on a larger scale. Preparation of 2-X-Imidazoles (3). 2-10doimidazole~~ (3a) was prepared from 2a by refluxing a solution of 350 mg (0.80 mmol) of 2a in 5 mL of 5% acetic acid in methanol for 30 min. After evaporation of the solvent, water was added to the residue. After chilling, the solution was filtered and the filtrate evaporated to give 155 mg of
crystalline 3a (99Oh),the homogeneity of which was demonstrated by thin-layer chromatography and chemical ionization mass spectrometry. Variations in hydrolysis condition are given in Table 11. The course of the reaction in each case was monitored by silica gel thin-layer chromatography. Registry No.-1,67478-52-8; 1-tritylimidazole, 15469-97-3. References and Notes (1) K. L. Kirk and L. A. Cohen, ACS Symp., 28, 23 (1976);C. B. Klee. L. E. La John, K. L. Kirk, and L. A. Cohen, Biochem. Biophys. Res. Commun., 75, 674 (1977). (2) C. R. Creveling, K. L. Kirk, and B. Highman, Res. Commun. Chem. Path. Pharmacol.. 16, 507 (1977);E. De Clercq, A. Billiau, V . G. Edy. K. L. Kirk, and L. A. Cohen, Biochem. Biophys. Res. Commun., in press. (3) 4-lodohistidine, initially identified as 2-iodohistidine,has been described previously: K. J. Brunings, J. Am. Chem. SOC., 69, 205 (1947).Later work corrected t h e structural assignment: H. B. Bensusan and M. S.R. Naidu, Biochemistry, 6, 12 (1967); M. S. R. Naidu and H. B. Bensusan, J, Org. Chem., 33, 1307 (1968).The synthesis of 4-chioro and 4-bromohistidine will be reported elsewhere. (4) K. L. Kirk, W. Nagai, and L. A. Cohen. J. Am. Chem. Soc., 95, 8389 (1973). (5) Y. Takeuchi, K. L. Kirk, and L. A. Cohen, manuscript in preparation. (6) 2-Chloroirnidazole has been prepared from 2-imidazolone:J. L. Imbach, R. Jacquier, and A. Romane, J. Heterocycl. Chem., 4, 451 (1967). 1Methyl-2-bromoimidazole has been prepared in poor yield from l-methylimidazole by reaction with cyanogen bromide: W. J. Langenbeck, J. PraM. Chem., Ser. 2. 119, 77 (1928). (7) D. A. Shirley and P. W. Alley, J. Am. Chem. Soc., 79. 4922 (1957).
J . Org. Chem., Vol. 43, No. 22, 1978 4383
Notes (8)Reaction of 1-methyl-2-lithioimidazolewith dinitrogen tetroxide yields 1methyl-2-nitroimidazole: B. A. Tertov, V. V. Burykin, and A. S. Morkovnik, Chem. Absfr., 81, 169 542 (1974). Iodination gives P-iodo-l-methylimidazole: B. A. Tertov, V. V. Burykin, A. S. Morkovnik, and V. V. Bessanov, Khim. Geterotsikl. Soedin.. 1109 (1973): Chem. Absfr.. 79, 126 395 (1973). (9) R . Jones, J. Am. Chem. Soc., 71, 363 (1949). (10) H. Giesemann and G. Halsche, Chem. Ber., 92, 92 (1959). A modified synthetic procedure was used, based on that for the preparation of toluenesulfonylimidazole: G. S. Schmir, w. M. Jones, and L. A. Cohen, Biochemistry, 4, 539 (1965). (11) For entries d, e, f, and g, chromatography was omitted in the workup. (12) B. A. Tertov, V. V. Burykin, and A. V. Koblik, Khim. Geterofsikl.Soedin., 1552 (1972); Chem. Abstr., 78, 58 308 (1973). (13) H. Gilman and S. Avakian, J. Am. Chem. Soc., 68, 1514 (1946). (14) D. I. C. Scopes, A. F. Kluge, and J. A. Edwards, J. Org. Chem., 42, 376 ( 1977). (15) T. Sheradsky, J. Heterocycl. Chem., 4, 413 (1967). (16) The material originally identified as 2-iodoimidazole [H. Pauly and E. Arauner, J. Prakt. Chem., 118,33 (1928)] wasshown \ater to be 4-iodoimidazoie (ref 3). Authentic 2-iodoimidazole has been obtained in 5% yield by reaction of iodine with 1-benzenesulfonyl-2-lithiolmidazole [R. J. Sundberg, J. Heterocycl. Chem., 14, 517 (1977)]; this N-protectinggroup was found unsuitable for general use in the preparationof 2-X-imidazoles, a conclusion we had also reached from early studies.
R
hi
H
I-I 12
9
13
R
H
I
H
14a, R = E t ; R' = C 0 , M e
15
b, R = C 0 , M e ; R' = Et
Reactivity of Oxoindole-A3~~-acrylates toward Diazoalkanes: An Unusual Ring Expansion Gregory B. Bennett,* Robert B. Mason, and Michael J. Shapiro
Department of Medica1 Chemistry Pharmaceutical Division, Sandoz, Inc., East Hanover, New Jersey 07936 Received April 4, 1978
As part of our work on the regiospecific behavior of enedicarbonyl compoundsla we decided to examine the reaction of (1). With diazodiazoalkaneslb with oxoindol-k7*1v-acrylates methane acrylate 1 (X = H ) 2provided pyrazoline 3a (X = H). The observed NMR coupling of the pyrazoline methylene and methine protons was sufficient evidence to assign structure 3a and not 2a to the product. Steric (tertiary vs. secondary carbon) as well as electronic (polarization of the C-C double bond) effects are such that the [1,3]-dipolar addition of diazomethane involves initial C-C bond formation a to the ester, with the reaction proceeding via a nonsynchronous intermediate such as 4 and not 5.3 When heated above its melting point or in refluxing xylene, pyrazoline 3a (X = H ) underwent Nz loss giving spirocyclopropane 7a (X = H). Reaction of acrylate 1, X = H, with phenyldiazomethane provided the corresponding spirocyclopropane 7b (X = H) as a single diastereomer. Exposure of acrylate 1 (X = CN)6to diazomethane did not afford either a pyrazoline (2 or 3) or a spirocyclopropane (7). Instead, only quinolone 1la (X = CN) could be isolated (92%). Rearrangement of the intermediate resulting from loss of N2, 8, and isomerization of the resulting exocyclic double bond out of conjugation with the cyanoester and into aromatization would account for the observed product.' T h e addition of a cyano group8 has thus reversed the polarization of the C-C double bond while equalizing the steric effects of substitution such that the diazomethane addition now involves initial C-C bond formation /3 to the ester moiety (5). The rearrangement of isatins to quinolones has precedent in the literature.9 Eistert and coworkers had reported that the reaction of isatin and N-methylisatin with diazoalkanes (RCHNz) led in good yield to 4-R-substituted 3-hydroxycarbostyrils (12). An equilibrium mixture of l l a (X = CN) and tricyclic 13 was established after only 12 h in MeZSO at ambient temperature. Dissolution of either l l a or 13 in MezSO resulted in the same mixture. The chemical shifts for compounds l l a (X = CN) and 13 were consistent with structural assign0022-326317811943-4383$01.00/0
ments made on the basis of other spectral data.1° The NMR spectrum of 1 la displayed a single aliphatic, methine carbon which disappeared on isomerization in Me2SO to 13. In addition the A ring aromatic carbons $ to N shifted downfield on isomerization, an indication of the imino ether tautomeric form. The rather high-field (80.2 ppm) absorption of the furan ring C n to the ester in 13 is consistent with a highelectron density resulting from mesomeric 0 and NH2 participation. Furthermore, hydrolysis of l l a (X = CN) followed by decarboxylation provided the known quinolone-3-acetic acid ( l l a , X = H ) , identical in all respectslla with the compound prepared by literature techniques.llb In a similar manner, reaction of 1 (X = CN) with PhC"iJ2 provided the corresponding rearrangement product 1 lb (X = CN) mp >325 "C in a yield of 42%.Reaction of 14a9C with diazomethane, on the other hand, provided pyrazoline 9 resulting from C-C bond formation a to the ester. When heated in refluxing xylene this pyrazoline underwent smooth conversion to spirocyclopropane 15. The double bond carbons of compound 14a are more sterically equivalent (tertiary vs. tertiary) than in the case where X = H and yet initial C-C bond formation has still occurred cy to the ester. This result lends support to the argument that only when C-C double bond polarization of the oxoindol-A,'."-acrylates has been reversed (as in the case where X = CN)13such that dipolar species react initially a t the carbon @ to the ester will rearrangement to the quinolone ring system occur. Furthermore, the sequence provides an efficient procedure for the synthesis of auinolone-3-acetates. Additional work relating to this rearrangement and the regioselectivity of such dipolar addition reactions is now in process. I
Experimental Section The IR spectra were recorded on a Perkin-Elmer Model 257 or 457 grating spectrophotometer and NMR spectra were recorded using either a Varian T-60 or EM-360 spectrometer. 'TNMR spectra were recorded using a Varian XLFT-100 spectrometer. Chemical shifts (6) are recorded relative to Me4Si; coupling constants (4are given in hertz. Mass spectra were recorded using either an LKB 9000 or an AEI MS-30-D5-50 spectrometer. Melting points were obtained on a Thomas-Hoover capillary melting point apparatus and are uncorrected. In all workup procedures, the drying process involved swirling over MgS04and filtering prior to evaporation. Ethyl 4',~'-Dihydro-2-oxospiro(3H-indole-3,3'-pyrazole)-4'carboxylate (3a, X = H). To a solution of acrylate l 2 ( X = H)(21.9 g, 0.1 mol) in anhydrous Et20 (700 mL) at 0 O C was added CH2N2 (ca. 5.1 g, 0.12 mol) (from 36 g of D i a ~ a l d )After . ~ ~ an additional 18 h at ambient temperature, the excess CH2N2 was quenched with HOAc and the solution was washed with aqueous NaHCO:?, dried, and
c 1978 American Chemical Society